Diyala Journal

of Engineering
Sciences

ISSN 1999-8716
Printed in Iraq

Vol. 04, No. 02 , pp. 118-129 , December 2011

BEARING CAPACITY BASED ON SPT-COMPUTER

INTERPOLATION
Dr. Khattab Saleem AbdulRazzaq
Lecturer

Dr. Waad AbdulSattar Hussein

Lecturer

Ali Hussein Hameed

Ass. lecturer

University of Deiyala\ Civil Engineering Department

(Received: 22/2/2010 ; Accepted:11/10/2011)

ABSTRACT:- Any structural design must be accompanied with sound analysis referring
to the foundation design. The columns carrying the total load of the building may be in very
high stress. The actual stress that exists in columns may reach in actual cases to half f'c or
more. On the other hand, the maximum carrying stress of soil is very much small compared
with that for reinforced concrete, the situation that necessitate the enlargement of column end
to have "a footing". While the compressive strength of concrete is easy to measure, the
bearing stress of soil is not. Methods of evaluating the soil bearing capacity are numerous and
consist of field and laboratory. The SPT is one of these field methods. Scientists tried to
relate the SPT-N value with the soil strength properties resulting in large number of tables,
charts, and graphs. This research considers the most famous methods to evaluate the bearing
capacity from the SPT. A BASIC computer program is written to aid in using these formulas.
In going to this step all tables, curves, and graphs must be converted to numerical equations.
This is done by using the usual FD technique of interpolation. The authors feel that this
program must be used with caution since it is not a replacement of sound hand calculations
associated with engineering judgment and experience. This is because the very SPT is used
only as a guide and never as a replacement of laboratory testing program except for sands
since it is very difficult to get undisturbed samples.
Keywords: Bearing capacity, Interpolation, structural.

LIST OF SYMBOLS
b
moist
s

vo
B
cu
Df
dw
Kp
N, N'
Nc, Nq, K

submerged unit weight of soil.

moist unit weight of soil.
saturated unit weight of soil.
effective angle of internal friction for soil particles.
soil settlement.
initial effective overburden pressure of soil.
width of footing.
undrained cohesion of clay.
depth of footing.
depth of water table = zw.
Terzaghi bearing capacity coefficient used in the formula of K.
corrected and uncorrected value of SPT.
bearing capacity factors based on .

118

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

PI
qall, qult
Rw, Rw'

plasticity index of soil = LL-PL.

allowable and ultimate bearing stress of soil.
SPT corrections based of A.R.E.A. for the presence of water table.

INTRODUCTION
The bearing capacity is a criterion for structural stability. Any structure, unless it floats,
must eventually be founded on soil. The failure criterion for foundation soil is known as the
ultimate bearing capacity or simply the bearing capacity of soil and is considered as one of
the corner stones of soil mechanics. For such a purpose, scientists from about many decades
ago tried to establish sound bearing capacity equations, which take into account the most
variables encountered. Nowadays, the available bearing capacity equations are "how we say"
numerous. Some of them have succeeded to float on surface while others have not.
By bearing capacity equations, here, authors mean, as well, all techniques used in field and
laboratory to "estimate" the ultimate bearing stress of soil. Most of the field data available are
presented as tables with boundary limits or nomograghs in terms of well standard tests such
as the SPT (standard penetration test) and the CPT (cone penetration test). The bearing
capacity equations based of theoretical approaches and laboratory tests are, of most, consist
of equations or nomograghs. The huge data, as a background, available for the SPT have kept
the light focused on such field test and instead of canceling it from soil investigation record;
it is still floating of surface till now.
The research has taken into account the most famous and well-proven tables and graphs for
field and theory bearing capacity estimations. Then a BASIC program has been established to
facilitate the use of these bearing capacity equations. For programming purposes, tables and
nomographs have been transformed into equations using the FD with most methods found
suitable for each table and graph. In carrying out this step, personal experiences have been
used based on authors' background. The aspects of the program are presented with brief
explanation for the procedures used and the FD interpolation equations reached. Some of the
source tables or data are presented in Engineering System. All data are transformed into Nkg-metric system.

ONE STEP FURTHER

To schedule programming the following scheme is adopted, see figure (1). The first
program page consists of many ways of estimating the bearing capacity from the SPT. These
methods are numbered and the user has to choose the method he wishes. The program will
require the parameters needed in the particular method if order to compute the bearing
capacity. Here, and in order to illustrate the program, each method is listed individually with
brief discussion about it. Some methods consist of direct application of a series of equations
leading directly to bearing capacity of soil. On the other hand, other methods consist of tables
and graphs. Moreover, as mentioned before, using FD interpolation the latter are transformed
into equations for use in programming. It must be mentioned here that in FD theory the closer
the value to the pivotal point the less error we get in interpolation, also the higher the degree
of FD polynomial the less error we get as well. We say in the outset that the pivotal points,
degree of FD polynomial, the FD method, are chosen in accordance to authors experience, to
the nature of point, soil type, numerical distance between points and so on. Any change in
one of those parameters will, in sense, change the FD polynomial. Nevertheless, authors
believe that most of these changes may be of minor effect on the value of bearing capacity.

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
119

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

FIRST PAGE CHOICES

The following choices will appear on running the program. A brief discussion will follow
with the number of case is referred:
1- Finite differences interpolation of submerged and moist unit weight, and angle of
friction- after Terzaghi and Peck- for sandy soils.
2- Finite differences interpolation of unconfined compressive strength of clayey soilsafter Terzaghi and Peck.
3- Allowable bearing stress based on the ultimate bearing capacity of sandy soils- after
Teng.
4- Allowable bearing stress based on one inch (25 mm) of settlement of sands- after
Terzaghi and Peck.
5- Allowable bearing stress for clayey soils- after Terzaghi and Peck.
6- Bearing capacity factors Nq and N obtained directly from the SPT-N value- with the
allowance of local shear failure in foundation soil- after Peck, Hansen, and
Thornburn- for sandy soils
7- Meryerhof equations for one inch (25 mm) or any settlement of sands.
8- Finite differences interpolation of angle of friction, and moist unit weight of sandsafter Bowles.
9- Theoretical Hansen bearing capacity equations- "for comparison", with Kenny
equation for very plastic soils.
10- It is usually costumed to refer to bearing capacity method by its scent's name or
names. The number between brackets show the reference from which bearing capacity
method is taken.

1-FD INTERPOLATION OF b, moist, AND FOR STANDY SOILS

This method is based originally on the empirical tables presented by Terzaghi and Peck
(1948), received many modifications later on, relating the SPT-N value with the relative
density, the angle of friction, and the unit weight of sands (submerged and moist). The N
values are corrected in accordance to the effective overburden pressure by a graph after Gibbs
and Holtz (1957). No mention to N correction for the presence of water table in bore-hole. On
the other hand, Terzaghi and Peck suggested increasing the angle of friction by 5 degrees for
soils containing 5% of fine sands of silts. The foregoing suggestion is not incorporated in
program for factor of safety.
The table for Terzaghi and Peck is transformed into numerical equations using the FD
interpolation. The relationships between N and b, N and moist are obtained by using Newton
divided FD, pivot N is selected as 10 degrees, while the equation relating N and is obtained
by direct linear fitting. The equations obtained are:

b
kN / m 3
(1)

N
moist

kN / m3
11.5

( 2)

0.28 N 27

(3)

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
120

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

The correction for depth is N

50 N
,
zw
10
6.895

where:
zw : depth of water table from the natural ground level.
N' : uncorrected N-value.
: moist unit weight or submerged unit, depending on the level of water table at
the time of the SPT-test.
zw
Two conditions restrict the depth factor equation, namely, 1)
must not exceed 40,
6.895
and 2) if N>2N' then the corrected N must be divided by a safety factor of 2. In practice, the
authors believe that this depth correction should be treated with caution since high values can
be obtained.
The FD fitting of b and moist do not match for N=10 and smaller with the original table
for Terzaghi and Peck by an error of 25%. This is not a serious problem since the angle of
friction for pure sand does exceed 26.5. This is called the "particle-to-particle friction angle
or .
And in order to find the bearing capacity of the sandy soil the unit weights and angle of
friction are used in Hansen equations to obtain the plain strain case of loading. It is worth to
mention that no shape or other factors are used in the Hansen equations since the TerzaghiPeck tables are considered crude.
It should be mentioned here that in case of the presence of water table in the DF range, the
soil water must be drained, by pumping for instance, for the purpose of concrete casting of
foundations. This ground water should never be used as a substitute for mixing water
according to ACI 31808 (3.4.1 and 3.4.3). On other hand, the values of qall or qult should be
compared with actual soil pressure under footing with ample safety factor as in ACI 31808
(15.10.3).

2-FD

INTERPOLATION

OF

UNCONFINED

COMPRESSIVE

STRENGTH OF CLAYS cu
Terzaghi and Peck presented a table similar to case(1) before but relating the SPT-number
with the confined compressive strength and with the saturated unit weight for clays. Linear
fitting is used between the undrained strength and the SPT-N while the correlation between N
and moist is ignored by program because:
1- It has minor effect on bearing capacity and,
2- The relationships between the undrained strength and SPT-N is very unreliable as stated
by Terzaghi and Peck.
Thus the bearing capacity is based on the term (cu Nc) with cu=5.14. This step is towards
the safety factor and is positive. No depth or shape factors are used. In sense:
cu=5.985*N
in kN/m2 , linear fitting
And
qult=cu Nc = 5.14 * 5.985 * N in kN/m2
(4)
No correction for N is used and qult is the ultimate bearing capacity of soils.

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
121

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

3- qall BASED ON qult FOR SANDS

Here qall is the allowable bearing stress of soils. Teng (1962) presented two empirical
equations relating the SPT-N number with the bearing pressures of granular soils. Gibbs and
Holtz correction for effective overburden stress, and A.R.E.A. correction for the water table
level are used in these equations. The equations provides qult of soils and Teng suggests a
factor of safety not less than 3. The following simple steps illustrate the procedure of
calculation qult with the aid of figure (2).

ALGORITHM
Enter depth of WT from NGL
If dw Df -----then-----Rw'=0.5, Rw 1

Df dw
2 Df

1 dw Df
If dw > Df and dw<(Df+B)------then------ Rw
, Rw=1.0
2
2B
If dw (Df+B)------then-------Rw' = Rw = 1.0
Correct SPT-N for effective overburden stress by using Gibbs and Holtz (1957) graph
Then for plain strain (PS) loading;
qult = 0.15709 [2N2 B Rw+6 (100+N2) Df Rw'] for square footing, (5)
qult = 0.150[3N2 B Rw+5 (100+N2) Df Rw']

for PS loading

(6)

Again, a safety factor of more than 3 is recommended to get qall.

4- qall BASED ON INCH SETTLEMENT IN SANDS

Terzaghi and Peck presented two equations for allowance bearing pressure based on
settlement of 25 mm in sands. Same corrections used in (3) before are used here as well. The
equations are,
qall = 34.47 (N-3)[

( B 0.3) 2
] Rw'(1+Df/B)
2B

qall = 6.89(N-3) [

( B 0.3) 2
]
2B

in case of (1+Df/B) < 2, and (7)

in case of (1+Df/B) 2

_ _(8)

5- qall FOR CLYEY SOILS

Terzaghi and Peck presented a table between the SPT-N value versus the allowable
bearing pressures of square and plain strain loading for footing resting on clays. A safety
factor of three in incorporated in the table with large settlement expected- as stated by
Terzaghi. FD-Newton divided differences of interpolations are used, setting the pivotal N=11.
At low N values, polynomial errors of about 13% are encountered between table data and
polynomial. The FD equations are,

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
122

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

qall=16.9N-0.048N2-7.76

in(kN/m2 ) for square footing, and

qall=13.55N-12.88-0.053N2

in(kN/m2) for PS loading.

(9)
(10)

6-Nq AND N OBTAINED DIRECTLY FROM THE SPT IN SANDY

SOILS
Thornburn, Hansen, and Peck presented a nomograph relating the SPT-N value with , Nq,
and N, and allowing for local shear failure. Now using the FD-Everett method and setting
No=20 for sands. The nomograph is transformed into two polynomials and then the Hansen
bearing capacity equations are used to find qult. The equations are,
N2
Nq=N3+0.8834
(11)
200
N=N3-0.14N2+3.2837N-15
(12)
The usual Gibbs and Holtz depth-overburden correction is used on the SPT-N value.

7- MEYERHOF EQUATION FOR SETTLEMENT OF ONE INCH ON

SANDY SOILS
These simple equations presented by Meyerhof give the allowable bearing stress on sands
based on 25mm of settlement or any other settlement.
qall 0.47 N
qall 0.4 N ( B

in case of B1.2m
0.3 2
)
B

(13)
(14)

Where:
qall in kN/m2
is the settlement in mm.
The N-value is corrected for overburden stresses. Based on authors experience the
settlement is restricted to 50mm as a maximum limit since beyond this limit the building may
suffer large distresses.

8-FD INTERPOLATION OF AND moist FOR SANDY SOILS

Bowles (1982) presented a table similar to that for Terzaghi and Peck relating the N-value
with and moist for sands. FD-interpolation for the table with No=12 and using the FDNewton divided differences, results in the following two equations,
N2
26.2 0.626 N
- - - - - (15)
143
N2
moist 16.074 0.147 N
in kN / m 2
- - - - - (16)
500
Bowles incorporated two types of SPT corrections:
1- According to depth from Bazaraa, and as follows
In case of vo 75kN/m2 ------- N=4N' / (1+0.04 vo)

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
123

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

In case of vo 75kN/m2 ------- N=4N' / (3.25+0.01 vo)

Where,
vo
is the effective overburden stress at the level of the SPT test.
N, N' are the corrected and uncorrected SPT-value, respectively.
2- According to the presence of water table in the bore-hole in addition the existence of
fine sands or silts (producing a negative pore water pressure when the SPT arm is pulled upresulting in a false increase in the SPT-N value). In such a case;
N 15
N 15
for N' 15
(17)
2
Bowles did not include in his tables a relation between SPT-N and the submerged unit
weight. Thus, unlike the Terzaghi and Peck, the program will require the input of the
effective vertical stress at the level of the SPT-test, and moist is listed in program as a
comparison with the in-situ one.

BRIEFS COMMENTS ON TERZAGHI AND HANSEN BEARING

CAPACITY EQUATIONS
The bearing capacity equations, in general, have similar form, that is, cNc+qNq+0.5BN
for plain strain loading. The worldwide equations used by soil engineers are that which
belong to Terzaghi, Meyerhof, and Hansen. These equations have proven to be the best
among others and have great history and practice ever. The general plain strain equation
cNc+qNq+0.5BN is modified by each scientist by adding factors that have some effect of
the bearing capacity of soil, such as the CD or UU for cu and , shape and depth of footing,
presence of eccentricity, ground inclination, and so on.
In sense, the Karl Terzaghi equations are the most famous and have long history of
successful use, but for program applications the Kp factor of N was presented by Table(1).
Using the FD interpolation to simulate Kp by one and only one polynomial is not an accurate
task, it requires either:
1-The use of several polynomials, that is, to subdivide the large range of (from zero to
45) into subintervals and each one is treated with, say, Everett formula. The total combination
of these equations will scope to full range the "mathematical equation of Kp".
2-Or the use of one high degree polynomial, say nine, if all nodal point are to be
considered.
In Table (1), two polynomials are shown, their equations are,

2
K

2.523 42.667 285

p 1500 15.31

(a4thdegreepolynomial
) - - - - - - - - - - - - - - (18)

Moreover,
Kp

5
43

3
2

0.9523
27.1467 393.32 2123 (a5thdegreepolynomial
) - - - - - - - - (19)
375000 60.483

because the Newton Forward FD is used, with initial o is 20, the points before o do not
match (ever) with the Terzaghi Kp coefficient. A more practical choice to the Terzaghi
equations for programming is the Meyerhof or the Hansen equations which, as well, have a
very successful record in foundation engineering with the advantages of the presence of many
factors that take into account many situations with affect the bearing capacity. The Hansen

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
124

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

bearing capacity equations are programmed as the last step in the first page, so that one in
concern can calculate the bearing capacity based on theory as for comparison with any SPT
past result. The Hansen equations are used without any correction factors (for plain strain
condition).
For very plastic soils, the determination of effective for drained condition analysis is
very uncertain due to the long period of triaxial tests needed. Kenney (1959), [5] provided an
empirical chart relating the plasticity index PI of clay with the sine of drained angle of
friction. It has been found that the linear logarithm equation will fit well. The equation is,
sin 1[0.806 0.229 log( PI )]
in raidians, and ,
- - - - - (20)

sin 1[0.806 0.229 log( PI )]( / 180) in deg rees

- - - - - (21)
Which can be used directly to find the bearing capacity for plastic soils in drained
conditions. In determining the bearing capacity using Hansen equations the following
approach is followed;
1- If the water table level is within the depth of footing, zw Df then the soil is assumed
to be fully saturated.
2- If the water table level is Df zw (Df+B) the water table level is assumed to be at the
footing level, zw=Df.
3- If the water table level is below (Df+B) or zw(Df+B) then the presence of porewater-pressure in soil is ignored.
These assumptions are considered in the program, and regarded on the safe side of design.

CONCLUSIONS
1- The bearing capacity of soils is a very difficult and complicated problem because of
the so many variables involved in the soil strength parameters and in the loading
conditions. As a sequence, the concept and derivation of the ultimate bearing stress
that a soil can withstand, differ from one scientist to another leading to so many
equations, nomograghs, tables, and so on. Because of that having one single program
to calculate the bearing capacity is an impossible task at least in the recent times. The
program in this research is considered rather simple but is collective for many theories
and self-experience.
2- The program is based fundamentally on the FD approximation to interpolate
polynomials. The FD method, the pivot points and the degree of polynomial are
considered as self-experience.
3- The program is mainly useful for office routine works, since in many situations risk
analysis is considered especially for the level of the water table.
4- As stated in earlier paragraphs, this program is not a substitute for actual design and
laboratory works, since the SPT is used only as a guide. Thus, the SPT is not a
substitute for the actual site investigation work as well.
5- It is rather difficult to differentiate between the methods as which one has more
accuracy since each method has it own assumptions.

REFERENCES
1- "Building Code Requirements for Structural Concrete" (ACI 318-08). American
Concrete Institute, Farmington Hills, 2008.

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
125

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

2- Bowles, J. E., (1982), "Foundations Analysis and Design", third edition, McGraw-Hill
Book Co.
3 - Craig, R. F., (2004), "Craig's Soil Mechanics", seventh edition, Spon Press, London.
4- Kreyszig, Erwin, (1983), "Advanced Engineering Mathematics", fifth editions, John
Wiley & Sons, Inc.
5- Lambe, T. W., and Whitman, R. V., (1979), "Soil Mechanics", John Wiley & Sons,
Inc.
6- Microsoft GW-BASIC, (1987), User's Guide and User's References, Microsoft
Corporation, U.S.A.
7- Teng, W. C., (1962), "Foundation Design", third edition, Prentice-Hall, Inc.
Englewood Cliffs, N.J.

Fig.(2): A.R.E.A. (American Railway Engineering

Association, Chicago, Illinois) .

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
126

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

Table (1): Comparison between actual Kp values and the FD-approximation.

Values of Kp

Terzaghi

Forth degree

Fifth degree

polynomial

polynomial

10.8

285

-2123

12.2

126.992

-776.005

10

14.7

51.98

-206.039

15

18.6

25.976

-17.132

20=o

25

24.992

24.682

25

35

35.042

34.365

30

52

52.137

50.877

35

82

82.289

80.173

40

141

141.52

138.204

45

298

255.818

293.919

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
127

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

Start

1st- Page Choices

from 1 to 9

On
1
On
2

FD interpolation of b, moist , of sands: Gibbs

& Holtz depth correction in addition to the use
of Hansen BC equations
FD interpolation of Cu in addition to the use of
undrained BC equations

On
3

qall based on qult or sands: use of Gibbs & Holtz

depth correction in addition to A.R.E.A. WT
corrections

On
4

qall based on 1-inch of settlement on sands

On
5
On
6

On
7

On
8
On
9

FD interpolation of SPT-N value to get qall of

clays PS and square foundations

FD interpolation of Nq and N directly obtained

from SPT-N for sands in addition to the use of
Gibbs depth correction
1-inch settlement
equations for sands

using

Meyerhof

FD interpolation of moist , , SPT-N for sands - use

of Bazaraa depth correction in addition to the use of
WT and silt pressure corrections Hansen BC
equations for qult

Use of Hansen equations

Drained
Analysis

Undrained
Analysis

BC

Use of Kenny curve

fitting equations for
Use of undrained Hansen
equations

End

Fig.(1): BASIC program flowchart.

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
128

BEARING CAPACITY BASED ON SPT-COMPUTER INTERPOLATION

)(SPT
.

superstructure substructure .
) ( .
.
. .
.
) ( _
_ _ .
) .The Standard Penetration Test (SPT

.
) (
. QBASIC .
. FD
.Interpolating Polynomials
Pivotal Points . .

Diyala Journal of Engineering Sciences, Vol. 04, No. 02, December 2011
129